Multi-Assistant Controller for Voice Command Conflict Resolution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current vehicle infotainment systems can only support operation of a single voice-activated assistant at a time, leading to conflicts when multiple assistants are enabled, as they may respond simultaneously to microphone inputs.
Innovation Solution
A multi-assistant controller with an audio recorder and detector that captures and processes audio data, identifies the highest probability wake-up phrase, and transfers it to the corresponding voice-activated assistant, allowing multiple assistants to be concurrently enabled and responsive.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple voice-activated assistants are enabled concurrently, then the system's versatility and user choice are improved, but conflicts and simultaneous responses occur
Solution Approach 1:
A centralized controller acts as an intermediary between the microphone array and multiple voice-activated assistants. The controller receives audio inputs, processes wake-up phrases, and selectively routes audio data to specific assistants based on detection results, preventing direct conflicts between assistants while enabling concurrent operation
Solution Approach 2:
The audio processing function is segmented into distinct phases: wake-up phrase detection (handled by the centralized controller) and voice command processing (handled by individual assistants). This segmentation allows multiple assistants to be active simultaneously without interfering with each other's wake-up detection
2Reliability
If a single default assistant is selected, then response conflicts are prevented, but user flexibility and assistant selection freedom are reduced
Solution Approach 1:
The system dynamically assigns default assistant status based on the detected wake-up phrase rather than using a fixed default. When a wake-up phrase is detected, the controller temporarily assigns that phrase's associated assistant as the active recipient for subsequent audio processing, allowing flexible assistant selection without predetermined defaults
3Speed
If all assistants listen to microphone inputs simultaneously, then response speed is improved, but harmful factors and conflicting responses increase
Solution Approach 1:
The centralized controller provides feedback to individual assistants about which wake-up phrases were detected and which assistants should process subsequent audio. This feedback mechanism enables fast response while preventing conflicts by informing assistants when they should remain silent based on the detected wake-up phrase
Data Source
AI summary
A multi-assistant controller includes an audio recorder and a detector. The audio recorder is configured to receive a sampled audio from a microphone, store the sampled audio in a circular buffer, and transfer the sampled audio from the circular buffer to a particular voice-activated assistant. The detector is configured to store multiple wake-up phrases that are recognizable by multiple voice-activated assistants, search the sampled audio to determine multiple probabilities that the sampled audio includes the wake-up phrases, select a particular wake-up phrase that has a highest probability among the probabilities, and send a callback to the particular voice-activated assistant that the particular wake-up phrase has been detected. The sampled audio that is transferred to the particular voice-activated assistant includes the particular wake-up phrase that was detected.